The number that hides a shrinking business
GMV and revenue can both be climbing while the business underneath quietly gets worse. Contribution margin per order is the number that catches it, and it is the number most SEA sellers do not track on any regular cadence, let alone weekly.
It is calculated as the price per order, minus cost of goods sold, minus platform commission, minus payment processing, minus the portion of logistics cost not covered by a platform subsidy, minus any ad spend directly attributed to that order. What is left is what the order actually contributed before fixed costs like warehouse rent, salaries, and software.
Why a monthly P&L review is too slow
A monthly review catches a margin problem after roughly four weeks of the same spend pattern already went out the door. Revenue growth driven by voucher-funded promotions or steep ad discounting can push GMV up while contribution margin per order on the exact SKUs generating that growth goes negative, and none of it shows up on a top-line revenue chart, because GMV counts the sale price before any fee or discount is netted out.
A seller watching only revenue and order count sees a good month. The same seller watching contribution margin per SKU sees a promotion quietly buying volume at a loss, three weeks before the monthly numbers would have shown it.
What moves the number most
Ad spend attribution is the term sellers most often get wrong, either leaving it out of the calculation entirely or averaging it across every order rather than attributing it to the specific campaign that drove a given sale. Attributing ad spend at the campaign level, not the account level, is what lets a weekly review catch margin erosion on one SKU while the rest of the catalog is healthy.
Platform commission is the second lever, and the one a seller has the least control over, since platforms set and periodically revise category rates. A rate that held margin steady at the start of a quarter can be a full percentage point different by a promotional cycle later in the same quarter, and that percentage point is often the entire margin on a discounted SKU.
Running it weekly instead of monthly
The mechanics do not need to be complicated. Pull price, cost of goods, commission, payment processing, logistics, and attributed ad spend for each SKU on each platform, once a week, in the same format every time. The value is in the consistency, not the sophistication: a sudden drop in one SKU’s contribution margin is visible the week it happens, not the month it happens.
Run it per SKU and per platform, not blended across the whole catalog. A catalog-level average can hold steady while one hero SKU’s margin collapses under a heavy promotional push, because the rest of the catalog’s healthier margin is masking it in the average. Breaking the number out by SKU and by platform is what surfaces the actual problem instead of a flattering blended figure.
Track contribution margin per order weekly, per SKU, per platform. Not monthly at the P&L level, where the promotions destroying it are already three weeks old by the time anyone sees the number.